Comment # 2
on bug 9427
from [email protected]
Guy,
Appreciate your analysis, agree mostly. Would say that all the payload is for
the conversations you found.
Further, we don't so much need full dissection of all layers of the traffic,
but just enough to lift out the what's in the payload, as if it were on an
Ethernet connection.
If further analysis is needed of what can be extracted/dissected, it can be
sent to the data provider. I will not be able to provide further, exahustive
information on the provider data.
How I analyze this data, upstream to the T3 source router, is to track the TCP
streams and retransmissions and match against any customer reported latency or
retransmissions, which we have seen are close together.
Personally, I belive it's either a telco issue, Cisco issue or spike in the
data. The last point is least weighty at present, due to the number of packets
in the queue, observed at about 50, and the limit is 90. We do see slips in
the telco (Guisys tool) and are working to analyze further. Oddly Cisco
doesn't see this! But can't say if failure to report is related to any
causality.
How can you proceed please?
Thanks - Brett
You are receiving this mail because:
- You are watching all bug changes.